[Subject Prev][Subject Next][Thread Prev][Thread Next][Subject Index][Thread Index]

Re: Experience the New Windows



On Thu, 22 Feb 2001, Atul Chitnis wrote:

->Eeps, methinks we are losing focus here.
->
->In general, Linux installs perfectly fine on most machines. I was actually
->trying to focvus the discussion on the post-install experience that gets
->users productive quickly, and familiarising them with what's available.
->
->Downplay the tech (assume pre-installed!) and focus on productivity and
->applications.

The post-install experience also depends on past experience and the
expectations carried from there.  It is legitimate to expect that the
documents made in one OS or office should be usable in others.  But
unless the applications adopt to open standards it is not possible.  This
is what I think is the major problem.  So I want to direct your
attention to what I think should be the focus, among others.

One of the main problems of offering a good linux desktop either at
home or office front, as we all know, is porting documents made from
applications made in windows to Linux.  The picture today is: sharing
a document made say in abiword, gnumeric, gimp, xfig, dia, TeX/LateX
to windows environment is not problematic. Because we now have good
RTF, ps, eps, html etc. filters available.  I have not experienced any
substantial information loss in this process.  The main problem
however is when we want to use or convert files generated in windows
environment to linux applications.  Sometimes we succeed and sometimes
we dont.  We often loose some information, particularly formatting
details, mismatching fonts, positions of figures etc.

Now the question is: Can we ever manage to solve this problem?  
I think, NO.  If our goal is to go `behind' the game and start
manufacturing better and better filters or emulaters then we are
never going to win, because by design we are always going to be left
behind.  

Let us look at the alternate scenario.  If we follow the open
standards, such as XML, for all kinds of applications, and make the
data transfer from linux to other environments easy and flawless, then
we throw the ball in the other court.  This I think should be the
strategy and our policy.  Sounds like advocacy again!  Cant help it.
IT needs a sound policy of information exchange, which is possible
only by adhering to global standards.

When someone sends us a `doc' document, we can request the sender to
send, at least in future, in open formats.  At least in the process
the word spreads that we are not using doc format here.  Many people
do not know that alternatives exist.

Meanwhile, we should also advocate (advocacy again!) why it is
fruitful to use structured document formating practices like TeX/LaTeX
for automating our information management.  It is not impossible to
make structured documents using office packages, in fact they should
be designed to do that.  In this regard I would like to share with all
of you a recent experience.  We recently had to convert one doc
document (a book of about 80 pages) to html, for obvious purpose of
publishing on the web.  No problems! We got one single fat html file,
which if published on the web directly without modifications, it would
have been an abuse of markup technology.  So we asked our project
assistant to do the necessary editing, and she took three days
improving and editing it.  On the other hand, I have converted another
very fat (400 page book) designed in LaTeX using latex2html filter.
It took 3 minutes on my desktop PC to generate the LaTeX code.
Without any editing I had a set of several html files, a stylesheet
file, some .pl files, and several gif files, with all the necessary
linkages made.  The original book has many cross-references,
footnotes, and also index at the end of the book.  You have to see to
believe it.  Everything worked as expected.  Though we needed not, we
have converted the same document into pdf and postscript versions of
that document.  That is the meaning of using computers for automated
document management and information transfer.

What I am trying to say by giving this example is, we should encourage
better IT practices by adopting open standards.  If we endup spending
our time in developing or trying out the various emulators, filters
for converting closed source information we are destined to loose the
game.

So what we need are applications that can generate structured
information by using powerful database, typesetting, and printing
languages in the background like, sql, TeX, postscript, and markup
languages like XML for data exchange.  Our future userfreindly office
packages should orient people by design into these practices, than
adhering to generating doc files and then developing filters for them.

To sum up: the problem of user friendliness is genuine, but the problem
of exchanging data across applications is an artificial creation
guided by selfish motives.

I didn't realize it is already got too long. Period!

Nagarjuna